Abstract Systemic lupus erythematosus (SLE) is characterized by the production of autoantibodies directed against self- antigens including DNA- and RNA-associated cellular components. Recent work from our lab and several other groups has shown that the nucleic acid sensing endosomal TLRs, TLR9 and TLR7, are required to break B cell self tolerance to these antigens and produce anti-DNA or anti-RNA autoantibodies. Despite these important insights, a great deal remains unknown about the origin of autoantibodies in lupus including the precise cellular differentiation routes that give rise to autoantibody-forming cells (AFCs). In particular, it remains unknown to what extent autoantibodies arise directly from the activation and differentiation of naive autoreactive B cells; whether an autoimmune B cell response can generate a pool of resting autoreactive memory B cells that can be reactivated over time; or whether both processes occur simultaneously. Recently a novel population of age-associated B cells (ABCs) was defined in mice and humans that are expanded in human SLE patients and in several murine models of lupus, including MRL/lpr. ABCs have been implicated as a candidate autoimmune memory B cell population and could be an important source of AFCs in lupus. These cells lack expression of B subset markers for follicular, marginal zone, transitional or B-1 lineages; express CD11c and CD11b; and express the transcription factor T-bet. ABCs have been postulated to contain TLR- dependent autoreactive and antiviral memory B cells, based on the presence of somatic mutations, expression of memory markers, and other properties. In this proposal, we will use genetic tools available in the MRL/lpr mouse model of lupus to investigate key questions about ABCs in vivo and to test the hypothesis that these cells are autoimmune memory and plasmablast precursors that promote lupus disease. In Aim 1, we will determine the lineage relationships of ABCs relative to naive B cells and plasmablasts, and determine the kinetics of their formation and activation. In Aim 2, we will test the hypothesis that ABCs are (or include) stable autoreactive memory B cells, and determine how the gene signature of MRL/lpr ABCs is similar to or different from that of conventional memory B cells or ABCs generated in normal aging. Finally in Aim 3, using genetic systems in which ABCs can be specifically targeted for constitutive or inducible depletion, we will test the hypothesis that ABCs are necessary for the initiation and propagation of lupus disease and evaluate their potential as a novel target for therapeutic intervention.